Metric structure from motion by indoor localization using Wi-Fi channel state information

Abstract

3D reconstruction methods for real-world environments based on camera images have recently been studied by many researchers and structure from motion (SFM) has been attracting attention as one of the most practical approaches for 3D reconstruction. SFM can estimate relative positions of cameras using many camera images and then reconstruct a 3D scene of the environment based on the positions. However, because the distances between cameras estimated by SFM have scale ambiguity, the scale ambiguity remains also in the reconstructed 3D scene. In this study, we propose a method for reconstructing a 3D scene with real-world scale in an indoor environment by estimating the absolute coordinates of cameras based on an indoor localization technique using Wi-Fi channel state information (CSI).